HBM2 isn’t being used in consumer products due to its high-cost. However, HBM(High Bandwidth Memory) is being used for professional grade GPUs and according to Samsung, the demand for the memory is so high that even if they double HBM2 production right now, they still won’t be able to keep up with the demand.
During the ISC 2018, Samsung noted that even if the company doubles the HBM2 production, they won’t be able to fulfill the market’s demands. This is another factor that has resulted in HBM2 being too expensive and a product with a high-profit margin for Samsung and SK Hynix.
However, the demand for this particular memory isn’t the only factor that has resulted in high-prices as there are a lot of benefits associated with the HBM2 memory.
For comparison sake, HBM2 can provide 8 GB of memory with up to 307GBps of speed on a single chip. This is something that even the Nvidia GTX 1070 can’t provide. Also, the single HBM2 chip requires only 1.2V.
Given the benefits, it doesn’t come as a surprise that even after doubling the HBM2 production, Samsung won’t be able to keep up with the market’s demand.
HBM being expensive is the major reason why consumer-level GPUs don’t make use of it, however, GDDR6 can offer performance relative to HBM2. GDDR6 has already gone into mass production and Micron has teased Up To 20 Gbps of memory bandwidth.
Micron’s GDDR6 memory overclock headroom showcases that upcoming Nvidia and AMD GPU featuring the GDRR6 memory might surpass the HBM2 bandwidth.
Furthermore, AMD has noted that it might ditch HBM2 in favor of GDDR6 for AMD Navi GPUs. According to AMD, while they have their eyes on HBM for its GPU but they are also planning to introduce the GDDR6 memory to the market.
It depends on how we see our roadmap, how we would like to play it out with some of our partners, and the innovations we want to have and what we want to do in the professional space. But we are fully committed to HBM and we’re going toWhatfully committed to GDDR6, and let the best solution win.
WHat do you think of HBM2 being highly in demand for professional-grade GPUs? Should GPU manufacturers feature it in consumer-level GPUs? Let us know in the comments.